Thanks for the info Tony.
It looks good in theory, but in reality that circuit does not work.
The Meyer waveform, I believe, should be channel A on, then both channels off, then channel B on, then both channels off, etc.
When breadboarded, your circuit only produces short spikes which occur at the same time.
Also, shouldn't there be a diode on the secondary of the output coil?
We have to start somewhere - I need to determine the relationship of the two complementary transistor driver circuit.
I think this may hold the key.
We can generate khz square waves quite easily with 555 timers.